1. Field of the Invention
Aspects of the present invention concern a platform to integrate data from multiple systems including, but not limited to, POS, video, EAS, alarm systems. etc. Some embodiments of the present invention relate to digital video recording and surveillance systems. More specifically, some embodiments of this invention relate to automated correlation of video data from digital recording video recording and surveillance systems with asynchronous transactions data from one-to-many discrete inputs.
2. Discussion of the Related Art
In the retail environment, certain types or patterns of employee or customer behavior at the register can be closely associated with fraud, theft or insufficient training. In addition, customer behavior at the resister is invaluable data for merchandizing and marketing. These types of behavior generally result in a monetary loss by the owner of the retail business. Analyzing point of sale (POS) data and searching for questionable behavior is gaining popularity among large retail organizations but is typically not practiced by less sophisticated organizations. Many organizations outsource this analysis and refer to it as “Data Mining” or “Exception Reporting”. In-house or third party systems presently in place are somewhat cumbersome and inefficient. The process of collecting the data from a multitude of geographically diverse locations, analyzing the data according to rigid guidelines, producing a written report, and distributing the written report can take several days and be very expensive. Further, the distributed Exception Reports are not tied to video data captured by surveillance systems in the store, as will be discussed in greater detail below.
For example, the use of surveillance systems to record cash transactions for later review well known in the art. For example, in U.S. Pat. No. 4,337,482 issued to Coutta, can be understood to disclose a surveillance system that monitors and records transactions that occur at a number of cashier lanes. In Coutta, a single television camera, mounted on a rail, can be positioned to make a video recording of the transactions that occur at a single selected cashier lane. Coutta discloses that the digital transaction data from the cash register in the selected cashier lane is fed into a video character generator to provide a composite video picture in which an alphanumeric display of the transaction data overlays the video image of the transaction. Since a composite video image is generated with respect to only one cashier lane, it is usually possible to position the camera so that the alphanumeric overlay does not obscure a useful portion of the recorded video image. However, if a single camera is used to simultaneously record transactions occurring at a plurality of cashier lanes, it is likely that the alphanumeric overlay data will obscure important parts of the video image of at least one of the cashier lanes. This likelihood is further increased when a large number of parameters are displayed simultaneously for all of the cashier lanes.
U.S. Pat. No. 4,630,110, issued to Cotton et al., suggests a surveillance system that monitors and records a single lane using a plurality of video cameras. In one embodiment, the video image from four cameras may be combined wherein two of the cameras are focused on cash registers. Cotton et al. discloses that textual data can be displayed at the lower portion of the combined video picture.
Another surveillance system disclosed in U.S. Pat. No. 4,145,715, to Clever, is understood to generate two levels of surveillance records. The first level, generated by a tape recorder, contains a record of all transactions. The second level generated on the tape recorder contains only selected transactions. In Clever, transaction data such as the price and department number are input to a character generator. The character generator output is mixed with the video image to create a composite video frame. This composite video frame consists of alphanumeric transaction data which overlays the transaction video image. This composite video frame is then recorded by the video tape recorder onto video tape.
Although Clever discloses that a single camera can be used to scan several point-of-sale (POS) stations, the generated video image contains alphanumeric transaction data that is permanently overlaid on the video image. This is because the composite video frame is generated before recording on video tape. This overlay can degrade the clarity of the resultant video images if the transaction data is placed over the video image corresponding to a particular cashier lane. Alternatively, a portion of the video may be blacked out so that the transaction data can be more easily read when viewed at a later time on the monitor. In this instance, the blacked out portion is recorded over a portion of the image being recorded by the television camera. In this situation, the portion of the video image which was blacked out is lost forever.
As the devices that perform data entry (cash registers, data terminals, optical character readers, radio frequency readers, magnetic media readers, and many other like input devices. etc.) become more sophisticated, larger quantities of alphanumeric characters describing the transaction are generated. The clarity of the video image of a transaction is particularly important when the transaction lanes are outdoors and are recorded under varying light and weather conditions. The increase in information tends to clutter and obscure the composite video image. As the number of lanes being recorded increases, it becomes more difficult to overlay all of the alphanumeric transaction data at positions that will not obscure important parts of the video transaction image.
As suggested by Katz in U.S. Pat. No. 5,216,502, transaction data and video pictures associated with the transaction data are recorded separately and synchronously on media capable of storing a full motion video.
The references described above all require that the transaction data be available at the time that the behavior is being recorded. However, there are certain applications where this technology cannot be applied, for example, in situations where the point-of-sale system buffers the transaction data until the termination of the transaction, or at the termination of several transactions, or when an event is associated after the transaction has occurred. At the end of the transaction, the data is transmitted from the terminal to the host. Accordingly, the transaction information cannot be recorded synchronously with the video pictures of the transaction.